21 Feb SPC/SQC LESSONS FOR THE SMALL MOLD BUILDER
A proper Statistical Process Control (SPC) / Statistical Quality Control (SQC) system can catch errors early to ensure machining consistency of molds and mold bases.
Machining of mold bases and molds for plastic injection molding is an exacting art form requiring tighter machining tolerances than standard machine shops. Mold makers are expected to hold +/- 0.002 or tighter compared to +/- 0.005 for many standard machine shops. At the same time, mold shops must maintain competitive pricing with rapid delivery.
Maintaining consistently tight tolerances is an ongoing challenge for any shop. Small mold builders too often rely heavily on the accuracy of their machining centers to maintain tolerances. Inspection measurements, if done at all, are mainly conducted using a dial indicator on a stand which is then dragged across a granite plate to check parallelism while thickness is checked using a handheld micrometer (see photos below for typical processes).
CASE IN POINT
If there is one call a mold builder, or particularly a mold base supplier, does not want to receive, it is a call from a customer complaining that the mold or mold base is out of tolerance. Corrections at this point are nearly impossible other than to machine and ship another mold or mold base which can still result in a very unhappy customer. It is at this point that the small mold builder will begin to look for the cause of the out-of-tolerance condition and hopes that it has not caused too many of their products to be out-of-tolerance. Besides being time consuming, if it is discovered that a number of finished products on-the-shelf are out of tolerance, the cost of replacement or re-machining can be devastating for the small mold builder. Such a situation can be caused by something as simple as an axis of a machining center going out of tolerance for some unexplained reason. That may sound silly but any machinist knows it happens.
SOLVING THE PROBLEM
The best solution available today is an independent inspection measuring system tied to Statistical Process Control (SPC)/Statistical Quality Control (SQC) software. Such a system will not prevent one’s machining system from going out of tolerance but it is the best assurance that the shop will catch the machine drift before it goes out of tolerance, and assure that out-of-tolerance parts are not shipped to one’s customers.
However, that presents another problem for the small shop. Where does one find such a system? How much will it cost? Do we have the technical resources to properly implement such a system even if we can figure out how to afford it? Is it worth the investment or are we better off to just be more careful in the future to try to catch any such issues with the system we now have in place? These are questions each shop must answer for itself.
Once a shop determines that it is prudent to research and possibly invest in such a measurement -SPC/SQC system, the first thing is to establish investment guidelines for the shop. For instance, the system should be:
- User-friendly to the point that it gets used by the shop on an ongoing regular basis, even when it does not seem necessary because the production machines are steadily holding tolerance
- Big enough to handle the shop’s anticipated largest parts or at least determine how large the parts the shop wishes to inspect on a regular basis needs to be
- Readily programmable to incorporate a growing number of various parts
WHAT MEASUREMENT SYSTEMS TO CONSIDER
Currently there are a number of measuring systems available on the market, many incorporating newly developed and emerging technologies. As expected, some are more affordable than others. For a baseline cost analysis, I suggest seeking out a used standard single point probe or attached laser Coordinate Measuring Machine (CMM). More robust newer and emerging technologies include:
- Points from Focus – Camera Created 3D Scans (see May 2021 Issue of Moldmaking Technology magazine – page 38)
- Optical / vision scanner systems
- Laser digitizing / scanning equipment
- Hyperspectral blue laser systems
- Confocal chromatic sensors
- Robotic gage arm CMMs
For a list of current inspection/measurement technologies and company provider contacts, see the July 2021 Moldmaking Technology 2021 Technology Review and Sourcing Guide section on “Inspection/Measurement,” beginning on page 87.
With regard to the statistical process control software, there are a number of options available. In general, it must be remembered that SPC/SQC has been around for many years and covers a broad range of technologies and production processes from machining to medical procedures to vegetation production. Schools offer semester and year-long courses in SPC/SQC methodologies, some of which get very complex mathematically.
Overall, SPC/SQC seeks to achieve process stability while improving capability through the reduction in variability. It does this by monitoring how a process changes over time in order to detect trends leading to out-of-control situations. Below is a typical SPC output from a robotic coordinate measuring system.
Statistical chart provided by FARO; displays a typical measurement output trend analysis
The internet is replete with instructional videos on SPC/SQC as well as providing Excel-based software that one can import and configure for one’s own shop’s use. There are also commercial SPC/SQC software packages that may be more suitable and easier to use. It is important that the software output a control chart, known as an “R-Chart,” for each machine feature considered important, especially those that are repetitive operations. Note that an R-Chart is a statistical variance over time. For each R-Chart, the shop will need to specify upper and lower control limits. This must take into account any natural or normal random variation around a desired or normal mean value / measurement. The control limits should be set such that they are reached prior to the part going out of tolerance.
One note of caution with respect to using SPC/SQC systems. Be certain that what you are measuring is statistically significant. Particularly, if you are a “techie;” these systems will allow you to collect all kinds of irrelevant data that seems statistically significant when it is not. Design your SPC/SQC system to collect only relevant data and thereby keep your shop operating at peak efficiency within the bounds of assuring the quality of those parts being “shipped out the door.”
ABOUT THE AUTHOR
Don Shrader is currently the Vice President of American Quality Molds (AQM) (www.aqmolds.com). A former Air Force Master instructor pilot, Mr. Shrader spent many years after his military service as a mechanical automation manufacturing engineer, conceiving and designing automated manufacturing equipment for many different companies. Prior to his involvement in AQM, he spent over 25 years as a consultant to Alcoa, Inc.’s Defense Sector working with Alcoa technical personnel to develop and implement improved aluminum alloy solutions in military aircraft that reduced weight while improving performance. Aircraft ranged from the aging KC-135 and C-5 to the F-22 and the new F-35.